Marine Technology, Vol. 31, NO. 4. October 1994. pp. 296-304

The International Sailing Canoe: A Technical Review

Paul H. Miller ' and David L. Dillon2

The unique teatures of the International Sailing Canoe have intrigued naval architects and sailors for years. This paper highlights the historical design development, describes current trends. and presents results from finite element analyses of three popular conSlruction methods: cold·molded cedar. fiberglass, and carbon/epoxy. Comparisons are presented of the bending and torsional stillness, pitch gyradius, and factors of safety. In general. the carbon/epoxy exhibited !he best characteristics. followed by the cold·molded cedar and the fiberglass. The Tsal'Wu quadrahc lallure crilerion developed for laminated plates was used in Ihe faClor·ol·safety calculations Factors 01 safety correlated closely to empirical development.

'!'HE International Sailing Canoe (Ie) has fascinated sailors and naval architects for years. Although outside mainstream sailing, its unique design and reputation for speed have maintained its esoteric reputation. The noted yacht designer L. Francis Herreshoff once wrote, ''The sailing canoe is . one of the most interesting things that God has let man make" IIJ.3 This paper highlights the Ie's history, presents the current state of the art in IC technology, and compares structural ana lysis results based on current construction and analytica l methods.

The paper is divided in four major sections. The first de· scribes the IC and reviews some reasons why the IC interests naval architects. The second reviews historical IC develop· ment, emphasizing technical changes adopted in the quest for higher speeds. The third section presents current trends, and the fourth a technical evaluation of three typical con­struction methods. The different construction methods---eold­molded cedar, fiberglass sandwich construction, and carboni epoxy sandwich construction-were evaluated using finite element analysis for factors of safety, radius of gyration and stiffness. These materials are also used in other sailboat classes, and the results illustrate the differences in current small boat construction.

Desc ri ption

The IC (officially , the "International 10 Square Meter De<:ked Sailing Canoe") is a 17-ft monohull sloop raced sin­gle·handed. Figure I shows a modern Ie sailing on San Fran· cisco Bay. The principal dimensions of the Ie are listed in Table 1, and a modern IC's profile view is shown in Fig. 2.

The IC class is considered a developmental one-design class, where certain aspects are controlled, such as total sail area, hull shape. and mast height. Other aspects, such as building materials. deck configuration, type of rig, and foi l designs, are developmental [21. The boats are raced without handicapping. Other well-known classes considered develop­mental one·designs are the International 14, the C-Class

I Senior composites engineer, Marioomp, Costa Mesa and graduate student. Department of Naval Architecture and Ocean Engineering, University ofCalirorma at Berkeley; 1989-1992 Intemational Ca· noe North American Champion.

2 Senior structural engineer, Marioomp. Costa Mesa, Califomia. 3 Numbers in brackets designate References at end of paper. Presented at the March 4. 1992 meeting of the San Diego Section

of1'm: SocIE"TY OF NAVAL ARCHITECTS AND MARINE ENGINEERS.

Catamaran, and the International America's Cup Class. For comparison, examples of stricter one·design classes include the Laser and J124, and examples of handicap classes include the IMS and PHRF handicapping systems.

The general characteristics that interest the naval archi­tect are readily apparent: the large length-to-beam ratio (5:1), the sliding seat, the very low displacement-to-length ratio, and the high sail area·to·displacement ratio. These make the IC a relatively fast boat. Speeds of 15 knots are common in stronger winds, corresponding to a speed-to­length ratio of3.6. During the 1991 National Championships in San Francisco the lead canoes finished the IO-mile course in under 40 minutes, averaging over 15 knots. Another char­acteristic t hat interests the naval arch itect is the simplicity and limited restrictions of the class rules. This has allowt.>d the IC to continue its development as experimentation is not restricted.

Approximately 300 Ie's are actively sailed throughout the world, with active fleets in the U.S. (New England, Chesa­peake, California), the U.K., Sweden. Canada. Australia, and Germany. The most recent World Championships drew over 60 competitors to San Francisco in 1993.

Historical de\'elopment

The IC's developmental history is similar to many other racing sai lboat classes in that it has been driven by the sail­or's desire to increase boat speed and ease of handling within the class rules. As with most development classes, the cur­rent designs reflect the boundaries of the rules and not the limits of sailboat design . For example, the current seal ex­tension limitation is 5 ft from the gunwale, and undoubtedly the boat would be faster in a breeze if the seat extension were increased to 5.5 It. This point should be kept in mind when understanding IC development, as many of the improve­ments were a result of rule changes and many other potential improvements were either delayed or abandoned because of rule changes or limitations.

Barly canoe racing (1865 to 1890)

The credit for inventing the sailing canoe belongs to Brit­ish Army Captain John MacGregor. While recuperating from a railway accident in 1865 he designed a small boat for ex­ploring the coasts of England and Europe. Designed mainly for use with 8 double-padd le, the boat was also rigged with a sail for downwind courses. The popularity of his book, A Thousand Miles in a Rob Roy Calloe, inspired others to build

296 OCTOeER 1994 0025-331 619(13104-0296$00.4310 MARINE TECHNOLOGY

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE OCT 1994 2. REPORT TYPE

3. DATES COVERED 00-00-1994 to 00-00-1994

4. TITLE AND SUBTITLE The International Sailing Canoe: A Technical Review

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Maricomp,Costa Mesa,CA,92626-2842

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution unlimited

13. SUPPLEMENTARY NOTES

14. ABSTRACT

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT Same as

Report (SAR)

18. NUMBEROF PAGES

9

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18

I

.L

Flg. 1 Tne modem Inlemal,onai Canoe

canoes and take up cruising, which led to the founding of The Canoe Club (later the Royal Canoe Club) in 1866 (3].

As the popularity of paddling and sailing canoes grew, more "canoe clubs" were founded . Many of these were located in areas that encouraged a particular aspect of canoeing; pad· dling (double·paddle), cruising, or sail ing. In the latter cate­gory the New York Canoe Club (NYCC) was founded in 1871 to promote sailing races off Staten Island , and later City Is· land in New York Harbor. In 1880 the American Canoe As· sociation (ACA) was founded to unite the efforts of canoe clubs across the country. In 1885 the NYCC offered a trophy, the New York Cup, for perpetual international competition ([3J, p. 145). The New York Cup currently resides in the United Kingdom, having been won from the U.S. in 1993.

LOAILWL Beam (canoe hu\l) Seat exte~ion from gunwale Draft (board up) Draft (board down) Weight (hull) Displacement

(wlo crew) (w/crew)

Sail area (actual) DisplL ratio (wlcrew) Saldisp ra tio (w/crew)

OCToeER 1994

5.18 m (17 III 1.02 m (3 Il" in.) 1.52 m (5 III O.l\ m (4.3 in ,) 1.12 m (3 118 in.) 63 kg (138 lb)

88.5 kg Cl95 Ib) (appro:.: .) 165.6 kg (365 lb) (approx.) 10.6 m 2 (1l4 Il~) 33.2 35.7

-,

Ie

T

Flg. 2 Sail plan

This trophy has the distinction of being the second oldest international yachting trophy after the America's Cup.

The canoes of that time were lightly carvel planked oak, mahogany, or cedar. The sailor would sit inside the craft. and lean to windward to counter the sail forces. Figure 3 shows the sailplan of Kestrel, an example of an early cruising canoe. In 1886 Paul Butler introduced the sliding seat, which greatly increased the available righting moment and, hence, boat speed. Figure 4 shows an early sliding·seat canoe sailing in the Alameda Estuary of San FranciSCQ Bay in the late 1880's.

British and American canoes diverge (1890 to 1933)

The next 40 years saw both a divergence of the British and American sailing canoe designs, and a decrease in the popu· larity of canoeing. In general , the American boats developed along the 16 x 30 line [16 ft (4.9 m) long and 30 in. (0.8 m) beam], and the British along a different development rule

Fig. 3 Sail plan Ql Kes!tel. circa 1890

MARINE TECHNOLOGY 297

fig. 4 Flying Scud. an early sliding·seat canoe sa.hng on San Francisco Bay. CIrca 1890 (plloto cour1esy NatIOnal Maritime Museum. San F •• ne.S(:o)

encouraging longer boats with roughly 0.08 m (3 in.) greater beam and higher weight [about 18 kg (40 lbl]. The American boats had more sail area [10.3 m2 (1 11 ft.:l) versus 8.9 m2 (96 ft.2)], but retained the unstayed cat· ketch rig [maximum height of 4.9 m (16 ft.)], while the British adopted a sloop rig (no maximum height). Sliding seats on the American boats grew in length to 3.05 m (10 ft), the British banned their use, and full-length battens disappeared.

During this period very light wooden masts were devel­oped . These 14-to-IS-ft masts were made by spiral wrapping veneers which were held together by lacquer. The finished masts weighed as little as 6.4 kg (14 lb).

Figure 5 shows Mermaid, considered the ultimate develop­ment in the 16 x 30 class, as she with skipper Leo Friede won the New York Cup in 1914.

Various sailing canoes, including Kestrel (circa 1890), Bee (circa 1890), Argonaut (cirea 1910), Mermaid (ci rca 1913, 1923) and an example of a "Rob Roy" Canoe (circa 1880), are well preserved a t the Mystic Seaport Museum in Connecticut [41.

The "International" canoe is created (1933 to 1971)

In 1993 the British yacht designer Uffa Fox designed and built two canoes, East Anglian and Valiant, to comply with both the American and British rules, and challenged for the New York Cup. To get around the rule differences for rigs. a free· standing sloop rig using a solid fore-stay was developed. Although not as successful against the British canoes, Fox and Roger Quincy won the New York Cup, as weU as the American National Championship [5].

The outcome of the English success was a joint proposal to create an international developmental rule for sailing canoes which combined the best features of each country's canoes. These included the sliding seat, greater sail area and lighter weight of the American boats, and the stayed sloop rig, longer length and greater beam of the British boats. Figure 6

298 OCTOBER 1994

Fig. 5 Merm,I(J. 19U

( - ----_ &: 4 ·

-.-- 1 ;:t= .,...j .- f_~ I

n.u..I_ i •• "..,

Fig.1S Flymg F,sh . art early " Internat.onlt!"· canoe. Circa '936

shows Flying Fish (circa 1936), an early British boat de­signed to the International Rule [6]. Compared with the mod· ern IC hull, Flying Fish is the same length, but is 3 in. nar­rower, and the aft chine is not as pronounced.

In both England and America canoe development pro· gressed slowly. Different boats were faster in particular con­ditions, but no design was clearly superior in all (:(Inditions. Innovations continued, such as:

• easier steering controls, • return offull .length battens in the 1950's, • composite construction beginning in the 1960's, • better hardware, and • aft sliding seat carriages fo r downwind sailing.

The modem "International Canoe"

The growing popularity and ease of fiberglass construction allowed the class to adopt a one-design hull shape in 1971 to en(:(lurage "mass-production" of composite IC's. A design by Peter Nethercott was selected due to its good reputation in all conditions. Deck design, foil s. rigs, and construction tech­niques were left developmental.

Although adopted for international competition, the one­design hull shape is not required for competition in the United States. In 1990 a single-chine canoe designed by Bill

MARINE TECHMQlOGY

Beaver to the u.s. National Rule won the U.s, Sailing Canoe National Championship.

T edmologicai developmenL'l sbCf:- the adoptIOn of the one­ct'Sign hull have focused on rigs, foil designs unci hull COli.­

struc~ion, In 1984 Americans mtrodulXld carb<:m fib€r r.msts at tho World Chumpior.ships and took the top throo spots. D,;;.ggerboards wHh smaller sections and shorter choid lengths have- proven faster than the older centerboardil. ex­per:ments with rotating ngs cO'!lt1nuc, and carbon has four.d its way tutu the hulls and nther structUTe.

Staie of the art in sailing eano(' design

Moden Ie technolegy refled;; current inmds in yacht de­sign. This section presents the des;~'D ami construc:ion trends and the last ~<;ction pn;sents structut$.l analyses of current construction mt't~ocl".

Hull (,(mstruction

Rule 6(g} of the Ie Rules governs hull c@strurtior., and simply states, "There are no restrict.ior.s on the material or method ofc0-n!'trUction ofthe hull" liZ;, p, 151 Some limita­tions however do exist; a minimum weigl-:t of US kg (138 lb) wid·. a limIt ef 5 kg (11 Ib) of corrector wBlghts for the strip.peo hull, and enough buoyancy:.o SUpP,)rt 75 kg {165 lbl witl: the hull Hoodod. 'two practical ccnsiderations not in· cluded in the rule" m"e also con.~iderod. toughness and ce!'t. CU7ently canoe!'. 8!'t bud: with one of a combination of mid. molded wood canstruction, fiberglass (wit~ or without a corel, or a-civanred composite cm:structwn.

The following three subsectioT\~ briefly desenbe currel'.t practices for each me':.hod, Because IC'~ art: raMly "mass pro· outed:' few canoes are identical This characteristIC limits the following d""criptions to genera! practices wit!'! a ft'W exampie;;. Some of the economic CDnsidHations of euch method are compa:ed and iater the different methods arc compared analytically.

Wooden hull collstructw~·Durmg the first 100 years of sailing CdUDeIL hulls were made of WGod, In order tominimize drag, carvd and strip plankir.g were commotl. although some boat." were lapstrake planked. After the Second World War canDeS began to be molded. A nun:ber were hot-molded using RI.'Sorcmd glue, but With the inCl"Ca&)d availabilIty of l'tmm­U'l':lpemture·cure ('pi.ixy resinfl, (,old-molding becamt) mort" common, Today some co:d-rnulded boats arc produced in A:ts­traha, Bnta;r., Canada, and Japar., particularly :n areas w'here molds for compo~itt' (on»tnlction are not available.

Cold.mold;Lg inv(l: VeS the IUymg of eMn veneers of a Jig-ht­weight wood OVCI" a tcrnparary m«le mold. The venet't\! are typically held together by epox.v, although other glues may be used. In the Unih:o SUtt-es the l.(;(:-:nquc is best kr.own as t}.e WEST syst€'Yn deve\opt·d by the GOl:geon Brothers f7l. The finished product is (;. s:.,rong, :ighc mtM((Xjce struc!ure exhibiting high flexural stlffnf's.s and fatlgu .. proper-:ies. In classes where construction rilles prohibit "advanced wmpos~ ite:;;" (,typically carbon or KevIa:), cold-moldlt::g offers a Com· petitive alternative to fiberglas!' i".-'Cnstrud i·m.

The prhnary options iI; cold-molding ate the number and thIckness of each vem~r, the :'Ipe('te!; of wood. and thB veneeT orit:r.ta.tian relative to the boat's r.~)llter!Jr.e. In Ie conc;tnIc· tion the number of veneer plies ranges from 2 to 5, and t!le thickness of each veneer (:.'Oro L6 ta 3.2 mm {l/lf W Va in.}_ The resulting skin thickness is typlca~:y between 6,4 aLd 9.6 mm (\f~ to % incht In North A:nerka, \-Vestern Red or Port Orford cedar is co:nmonly used, although mahogany may be ;jse;; as thf~ outer I: ft)r ?-dditionai toUjithlle&S, and a layer of

60 to 200 gicm2 (1.75 to 6 oziyd:!) fiberglass doth may be adclod to the outside for abrasion resistance.

Rt:.cking sequeneCb rmege from ~90"iO~] (';;0 the fRnterlw£) for $: two-ply lam'.nare, :0 a more ¢cmpliC£<~ed Jamu:.ate in­c>Jd:ng off-axis paes of 45~ or 30", The advantages of ufling more plies are greater control of properties, and a general:y higher fatigue life due to the n:-duction 0: shear loading on the glue The dlsadvantages are B. r.igher weight for a given th)ekuoss dm; to thc addlt:'onal ghw, and more labor. The prefertmce and attua: superiority of one over tr:e othl]r is a function of the amount of effort and qualit:' the builder (an \-{iterate, Generally, ahoui 40 man-hours are required to lay a single layer of veneer on a canoe hull {personal eonnnur.:'ca­tiou with Fran Def'aymoreau, Feb. 19921.

Pibf'rglass hull COrl--5trucfion-The dosire to reduce con­s:ruction times (and costs for non-owner-built boats) inspired car.oe sailors to tr)' fiberglass cCllSU"uction. In the United States, Lou \v1ritman. who dominated American Ie develop­uwnt and sailing from the 1940's to thi' 1!)71}'s, built the first fiberglass boats in the illte 1960's, During the 1970's and 1980's, SteVI] Clark, Too Van DUG(ll), a:J.d Erich Chase furt::'er developed fhe fiberglass Ie. These boat<> were built using materials and methods ,,'OM\1Wn in othe.r small craft construe--­tion, and many Sre still competitively raced.

T:nncal fiberglass- construdlOO involves the ~?-turatlOn of tl dry' IE>gJassi fiberglass cloth with <\ polyeste:r resin, the plies being placed in a female meld. In "roaG where greater flox­ural stiffnt--&il and strength lire required, a core is u~ed be­tween the P~lCtl. In general, Eberglass lamirH\tkS are cured at room temp0rat.u:re and preSi<Urk. The options ill 5.berglass (.."(Instructwn ?-re similar t,o t'oJd-molded construCtl1JI'_: ply thickness, doth fottn;):t and weight, and ply orientation [8l A typical Ie Hberglass huH lay-up takes about 6 man-hours.

FIberg!ass Ame!lctln sailing canoes typically were buE: with between 850 to 1020 glcml (25 to 30 oz:yd2

) (dry) outer c;k"iTI laminates and 270 to 410 glCtn2 :8 to 12 ozJyd2) (ciry) innf'r sk!n laminates, wit!: C<:Irfl thickness ranging from 3.2 to 9.6 mm 'Y& to +'~ in.). The doth warp direc!:on was ger.erally laid perpendicular to the centerhne fo-r greatest economy, Mos': canoes wen! "composlte" amstruction, with fiberglass hunt' and plywood detks The scrvit.:e·lifc problems of these boats included insuffifient strength in tl1fr ply-wood decks and a "softening" with age due to the low fatigue rCSlstanee of polyeswr refin combint:d wi.th a light E-gIMS Jamina~e 191. Two fa..-wrs led to the usc of advanced composite construction in the Ie; the desire to b'..Ii1d hoats With longer con-.petitive Efe spans and 1ighwr "ends," and the goul to avoid many of the service Ete prablen1s of the wood and fiberglass boats.

Adcanced com.pDlilte construawn.-Thc term "advanecd composite" con!ltruction in the marine inciuht!)' generally re­fen; to the use of higher-modulus rcinforceme!\t fibcrs find

hig!tt'r perfoTnultlce resin SySu!ffiS thun the typical poly('ster.' E-gla<>.S combination ll!led in "fiberglaljs" construction. In some cases the tcnn also indudes unidirecti:::nal E-glass fab­rics, and some "advanted" constructidn techn;ques such as vacuum·bagg:ng.

Ie builders began using these sl"lvan,:ed materials and teehnlqucs in the early 198:0's im ¢omponents offering the greatest performance increase, Ttlese include-:: bnnded ,ar­bor. fiber masts developed by Ted Van Dusen and vacuum­baggod carbon fiber daggerboards by Steve Clark, The mate· rials were generally Mom·u·mperature cured epoxies oomhined with a 1'·300 grade carbon. Using raroonre-::uceci the !":Hl.st weight by 30~1

CarlxJn did no~ find its way in~o IG hull!; until tho late 1980's, American eanae buil:iers first began by orienting uni­-ciiredional rlies to match. the load directions. and tapering cores in the lower stress areas neal' the bow and stern, D'.lr­iog the mid-tu80's vacuum-fwg"ing-, to increase the ply com·

paction and reduce the resin content., and later post.-curing to increase the resin properties, became routine.

Economic comparl$on of hull materials- Table 2 compares estimated costs to build an IC using the three methods de­scribed above. As with any cost estimate, the assumptions drive the bottom line, but valid conclusions can be made by adjusting the assumptions to fit a particular case. In this case the actual material costs and labor hours were considered , with the assumption the builder had the necessary molds and tools available. The wood boat was assumed to consist of a three· layer hull , and the composite boats were five plies plus a core; decks were slightly lighter. In each case, costs for hardware, sails, mast and rigging were identical , and repre­sent current de!ligns.

The comparison shows that based on the material costs alone, the fiberglass and wood boats are approximately equal , but an all -carbon boat is approximately 65% higher. With an arbitrary labor rale of $lOlhr included, the ftber­glass boat is now 22% less than the wood boat, and the carbon boat is 20% higher than the wood boat. For a three-layer wood boat, the break-even point with a carbon boat would occur when the labor rate was $25 .50lhr , and a four-layer wood boat would break even at about $lS.50lhr. Figure 7 compares the elfects of labor rates on the total cost of a three­layer wood , a four-layer wood, a fiberglass, and an all-carbon IC .

A boat combining carbon and glass in the same laminate would fall between the fiberglass and all-carbon lines, and depending on the amount of carbon used and the labor rate, it may be more economical than a wood boat.

Tlble2 CO" Mil"'''" for dll'tw"'l con"NC1ion t.ehtMquu

Fiber- Carboni Material: W .... ,,~ Epoxy

5-ply 5-ply laminate; 3-layer w{c:ore wfrore

Labor. hull 120 30 60 dKk ' 0 30 ,0 hardware 20 20 20 daggerboanl 15 10 10 rudder 20 20 20 rigging 15 15 15 tot.allabor hours 230 125 165 labor rate per hour $10.00 $10.00 $10.00

Materials (hull. deck . &eat, foils) '

wood (fi l) 48. 115 115 resin (gal) ' .3 5.3 5.1 cloth (fl.~) 71 5 .25 825 wood ( $lft~) '090 $1.20 $1 .20 resin ($lgal ) &44 .00 $32.50 $44.00 cloth ($Ift2) $0.21 $0.36 $2.35 wood c:ost " ,. 50 $136.00 $138.00 resin eoet $1 89.20 $172.25 $224.40 cloth cost $1 5.32 $294.64 $1938.75 hardware C08t $475.00 &475.00 $475.00 8811 cost $650.00 ..... 00 ..... 00 mast and rigging rost $875.00 $875.00 1875.00

Total material cost: $2641 .02 $2604.89 $4301.I5 Total labor cost: $2300 00 $1250.00 $1650.00 Total (()&t: $4941.02 $3854.89 $5951.15

NOTES: 1 Man-hour and material e-stirnat.et based on dl8CU53ionll with cur-

rent canoe builde". 2. Material cost estimates from quotes obtained between Jan. 20

and Feb. 25, 1992, and do not include talles. 3. Labor rate for comparison purposes only.

300 OCTOBER 19904

11000

10000

900O

... 000

• 01000 u .. '000 • (; 5000

.. 4000

lOOO

2000 0 •

WOOD (.·LAYER)

WOOD Il· LAYER)

ALL·CARBON FIBERGLASS

,, - --" .. -

10 IS 20 LABOR RATE "

. -.­...

" Fl9- 7 Labot r'le versus total cost 01 an Inlerniit lOnal C,f\OII

Rig design and conslrudion

Adopting the one-design hull shape forced canoe designers to look elsewhere for performance improvements. One area receiving attention was the rig . Improvements were sought in aerodynamics and rig weight reduction. The reduced rig weight would improve boat speed in waves, and improve boat handling characteri stics. This subsection highlights the cur­rent state of the art in rig development.

Rig materials-The IC rules do not restrict materials used in mast or boom construction , and no minimum weight is required (12), p. 16). Masts weigh 10 to 20 lb when rigged. Current IC masts are made of wood (strip.planked or tradi­tional hollow construction), aluminum , or carbon (braided or hand lay-up). Each has Its advantages, and none seems to dominate the competition. Table 3 lists many of the trade-oil's in material selection. Note that carbon masts are typically less expensive than aluminum masts in the U.S. because carbon masts are manufactured domestically and the pre­ferred aluminum masts are made in the U.K.

Rig cUsign-The IC rules limit rig design in the following ways ([2), pp. 16, 17)1:

(a ) The thickness of a rotating mast shall not be less than Vl of the chord at the same position.

(b) A maximum mast height of approximately 6.25 m (20 n 6 in.) above the water.

(c) The total measured sail area (all sails) is 10.6 m2 (114 n2 ).

(d) The mainsail must be able to be lowered while on the water.

Current rigs generally fall into two categories. fixed or rotating, with fixed rigs preferred by about 95% of canoe sailors. The mllin advantages orthe aingle-spreader fixed rigs are lower weight, easier tuning in dilferent conditions, and a mature sail design to match the mast's bending characteris­tics. Mast sections are typically round with diameters rang­ing from 50 to 65 mm. Fixed rigs took the top three spots at the last World and National Championships.

Rotating rigs are still considered in the experimental stage, with wide variation in design concepts. Three rigs showing the most promise were the single-luff" aluminum mast rig developed by Chris Converse in the early 1980's , the double-lull'spruce-earbon mast developed by Paul Miller in the late 1980's, and the single-luff carbon rig developed by Erich Chase in 1990. All three rigs won races in light to moderate conditions, but did not perform well in strong winds. The best performance of the Converse rig was second place at the 1984 World Championships. The Miller rig won the 1989 North American Championships, and the Chase rig won one race at the 1990 World Championships.

The main disadvantages of the rotating Converse rig was the higher weight than the fixed rigs, which increased piU:h-

MARINE TECHNOLOGY

Mat.eriaV Characteristic Wood

Easily customized Inexpensive Amateur

construction

Lower properties Inconsistent

properties

Aluminum Carbon

Wide selection Relatively Consistent inel[pensive

properties Light

Heavy Limited selection

Cost (unrigged) $50-$200 (home built) $600-$950 $500-$650

ing in waves, and the lack of control over mast bend, which resulted in a loss of rig tension and pointing ability in strong winds. The Chase rig solved the first problem by using car­bon/epoxy in place of aluminum. Mast control remained a problem, however.

The Miller rig was designed with a larger cross-section for increased stitrness to maintain rig tension for the same weight as the aluminum masts. To otrset the increase in mast-diameter-induced drag, a double-luff mainsail was used. This rig type was originally designed and patented by L. Francis Herresholffor use on the R-boatLiue Yankee in the late 1920's, but the rule makers outlawed its use before it could be tested (111, p. 21). To maintain the same mainsail weight as the single-luff sails, a lightweight Mylar-film sail was used . This rig showed promise, but heavy weather per­formance was not equal to the fixed rigs due to difficulties controlling mast bend. Both the Chase and Miller rigs also su ffered from a lack of sail development.

Structural analysis of modern canoe const ruction

The three construction methods described earlier were compared using finite element analysis (FEA). The pw-pose was to determine what impact the added technology would have on the performance of the IC, and where modifications could be made to reduce weight in the ends of the boat.

Finite element analysis is a mathematical method of de­termining deflections and stresses in a structure. The struc­ture is broken up into smaller "elements," each of which is given stiffness and strength parameters based on shape, type of element, and material. A mathematical representation of the structure is then constructed in matrix fonn based on these stiffness parameters. This stiffness matrix is then used to solve for the resultant structure displacements by using the applied force vector and basic matrix algebra. Stresses in the clemcnts nrc found through back substitution using the stiffness matrix and solved displacement vector. Finite ele­ment analysis has been widely used in the aerospace and other industries since the 1950's, and its increase in use has paralleled that of the computer due to the computationally intensive nature of FEA [10].

Laminate

Hull fwd of mast Hull an of mast Chine Foredeck Main deck Bulkheads Kingplank

OCTOBER 1"

Wood Model

[-45/45/9010] \-45/45/90IOJ [ - 45/45f90/01 10l9OJ [0/90/90/0] [0f90190/0 I [0/90/90/01

These analyses were performed using COSMOSIM, a gen­eral-purpose finite element code run on a personal computer. Due to the simple design, a global model was built using composite shell, beam, truss, mass, and isotropic shell ele­ments [11 J. Every major structural component on a typicallC was modeled, including the sails, rig, foils, and sliding seat. The model was loaded using uniform pressures based on an upwind sailing condition applied to the sails, hull, and dag­gerboard, and was constrained at the aft end of the boat. These boundary C1lnditions do not produce actual vertical de­flections due to the imbalance of localized vertical forces, al­though relative vertical deflections occurring forward of the constraints are accurate.

The completed model size was 2602 nodes, 2942 elements, 15600 degreesoff'reedom, and took approximately two hours to complete the linear analysis. Each analysis required ap­proximately 60 MB of han:! disk space, and each model was evaluated in four areas: bending stiffness, midsection torsion, pitch gyradius, and factors of safety. The method of approach used in these analyses was based on previous work by the authors for Team Dennis Conner's 1992 America's Cup de­fense effort.

Model descri ptions

A standardized "typical" IC was used for all three models, so that the only differences would be in the structural mate­rials. Between models, different laminates were used in the fo redeck, main deck, hull, chine area, and bulkheads. These laminates are given in Table 4 for each model. The materials used in each model were typical values for those used in Ie construction, and the material properties used in the analy­ses are listed in Table 5. In general, a O-deg ply is considered to be fore/art, and the plies are listed from inside to outside. For bulkhead laminates, 0 deg is athwartship, and the stack­ing sequence is given from aft to forwan:!. A four-lnyer wooden boat was chosen for modeling due to the greater ef­ficiency of the four-layer stacking sequence.

The mast, boom, and forestay were modeled as beam ele· ments; the mast properties were based on a carbon fiber mast. The forestay was preloaded with a 250 lb force. The

FiberglaS!! Model

[0:210.125 in. core/0:3J [0:210.25 in. corelO:31 [0:51 [010.0625 in. corelOj [0145/0.25 in. core/45/01 [0/9010.5 in. core/90IOJ [019010.25 in. oorel9OlOJ

CarbonlEpol[Y Model

[0/45/0.125 in. corel014510J [0145/0.25 in. core/0/4510] [0/4510/45/01 [0/0.0625 in. core/O] [0/45/0.25 in. oore/45/0j [019010.5 in. core/90IOJ [0190/0.25 in. corel9OlOj

MARINE TECHNOl.OQV 301

T.bIe 5 .... ari.1 propertIH uMd In .n.lysls

Material: Cedar/Epoxy E·glass Cloth Carbon Cloth Uni E·glass Uni Carbon Core

Mfg. deaig. 181 282 6 oz. . ". 9 lblf\? ()zIyd~ (dry) 18.0 8 .9 4.7 6 .0 50 26.2

Property: E. {msil 1.6 30 9.2 5 .6 17 0.07 ? (mai) 0.075 3.0 9.2 0 .5 0 .5 0.07

., (msi) 0.075 05 0.9 0 .5 0 .9 0.03 NUrk 0.3 0.04 0.06 0 .27 0 .3 03 X, ( si) 8 50 85 135 200 0.15 X< (ksiJ 4 28 77 88 120 0.25 Y, (k5i) 0.25 47 83 45 8 0.15 Y< (lui) 0.4 28 " 17 14 0 .25 XY, (kei) I 1.8 19 8 11 .1 I X, (]blin. ) 500 400 765 743 1200 38 X< Oblin.) 250 224 693 484 720 63 Y,Oblin.) 16 376 747 25 48 38 Y < (Ib1in.) 25 224 666 94 84 63 ll~ (in.) 0 .0625 0.008 0.009 0.0055 0.006 0.25

Spec. densi ty 0.39 2.6 1.75 2 .6 1.75 0,\4 Fiber dens, (pei) 0.0 14 0.094 0.063 0 .094 0.063 0.005 Fiber voL 99% 57% 40% ,,% 64% 100% ResIn weight 5% ,,% 53% 28% 30% ... Ply wt (Ozlft2) 2.10 136 1.11 0 .93 0 .79 2.91 No. of plies in hull 4.00 5.00 5.00 1.00 Skin weight Obi 34.1 27.6 22.5 11 .8

Hull weight Ob} Wood Fiberglas.s Carbon 34.1 39.5 34.4

NOTE: These material properties are for analytical purpoeeq only and should not be used for deSIgn.

shrouds were modeled using truss elements to represent the end conditions of wire rigging used on an IC. Only the wind· ward shroud was modeled, as the leeward shroud is slack during upwind sailing. In all three models, a wooden seat and seat carriage were modeled, with a 79 kg (175 lb) mass (rep· resenting the sailor) placed on the outboard end . Figure 8 shows a typical undeformed and deformed model, as produced in these analyses.

Bending stiffness

Bending stiffness was evaluated by taking the vertical de· flections at the bow, shroud attachment, and mast step and finding a total deflection at the mast step by using the for· mula

Total 8 = (0 shroud + (8 bow - & shroud)

x 12.94/82.66) - & mast step

The model with the lowest overall deflection had the high. est bending stiffness. Bending stiffness is important as it af·

FIg. II TYPIcal undelormed and oelormed vie~

302 OCTOBER 19904

fects the rig deflections, sail shape, and hull shape; sti ffer is considered better. The overall deflections for each of the three baseline models are presented in Table 6, and a contour plot of resultant deflections is shown in Fig. 9.

The results show that the carbon/epoxy boat had the larg. est bending stiffness due to its superior material properties. The different laminates used between the fiberg lass and car­bon boats illustrates the impact of double bias plies to reduce resin shear loading and increase fatigue life. For the same laminate the carbon and fiberglass boats deflected in equal ratios to their respective EA. The carbon/epoxy modulus is roughly three t imes that of unidirectional fiberglass . For this study the overall deflection criteria indicated that the car· bon/epoxy boat with the mixed biaxial and double· bias lam· inate was 25.3% stiffer in the bow section than the wood IC, and 47% stiffer than the fiberglass boat.

Midsection torsion

Midsection torsion Wil l> evaluated at the mast step, the dag. gerboard centerline/shroud a ttachment, and at the forward end of the seat carriage. Torsional stiffness is important to maintain rig tension and rigldaggerboard alignment. Sec· tional torsion was calculated by taking two sets of opposing nodes at each section, and using their or iginal and deflected positions to define the rotation. The results from these cal· culations are given in Table 7, and a typical rotated section is shown in Fig. 10.

Model

Wood f iberglass Carbon/epoxy

it Mast it Bow, in. i1 Shroud, in. Step, in. Total d , in.

0.1199 0.1785 0,0922

0.0982 0.1362 0.0622

0.0332 0 .0463 0.0158

0 .0684 0._ 0.0511

MARINE TECHNOLOGY

l.~ .... 111

••• .. "'" .­••• •. n!» .. -••• '.RlU .. -••• .. -.. -il-U

FIg. II Contour plot 01 resultant deflectIons

T.bla 1 Hull tOf.1otIaI1 rotation II three MCtlon,

Model

Wood Fiberglass CarbonJepoxy

Rotation at Mast Step, deg

0.261 OA1 5 0.221

Rotation at Daggerboard .

d.g

0.423 0.640 0.311

Rotation at Forward End o(

Seat Carriage. deg

0.44 1 0.694 0.374

lJiDEFORMEDo.,..,F-7

Flg. l0 Typical rolaled seclion

As with bending stiffness, the carbon boat is stiffer torsion· ally. Here, the carbon/epoxy Ie is an average of 19% stiffer torsionally than the wood IC, and an average of 48% stiffer torsiona lly than the fiberglass IC, much of which is due to the off-axis plies.

Pitch gyradius

Pitch gyradius was calculated by the FEA program, and is presented in Table 8. All three models were adjusted with "corrector weights" so that the total weight of the boat would be the same. This method approximates the class rules. Therefore, since the rig weights were also identical, a lower calculated pitch gyradius indicates a hoat with a better dis­tribution of weight fore and aft. This is important to the sailor because a lower pitch gyradius will provide better sea­keeping for the boat, and is thought to improve performance.

The carbon/epoxy IC has the lowest pitch gyradius of the three models, although the differences are not great. The car­bon/epoxy IC model's gyradi us was 1.5% lower than the wooden boat FEA model's, and 1.8% less than that of the fiberglass boat FEA model.

Factors of safety

The factors of safety in the structures were determined using the Tsai· Wu fai lure criterion for laminated plates. This quadratic failure criterion produces an elliptical failure space which is more appropria te for laminated structures, where resin failu re is important, than a maximum stress or strain criterion that includes shear effects.

The Tsai-Wu failure space equation is:

OCTOBER 1994

Model Pilch Gyrsdius. in.

Wood Fiberglass CarbonJepoxy

(1IX, + lIXJ x (JI + (l IY, + l /Y~)

40.598 40.727 39.977

x (J2 - (J~/{XI x X <) - (J~(y, x Y) + TU,s'l

+ (2 X Fl:/ X (JI X (J2)/(X, x X~ x Y, x Y)

where

XI> Xoo Yh Y<, S = material strengths 010 (J2, TI2 '" material stresses

F i 2• == Tsai-Wu interaction term

If the value of the equation is less than one, the structure has not failed, but if it is one or higher, failure has OCCUlTed. This formula can be used to fi nd the factor of safety. One problem with the Tsai-Wu failure criterion is that no infor· mation is provided about the failure mode. Many engineers now use Hashin's failure criterion, which does provide infor ­mation about the predicted mode of failure, but this option was not available in COSMOSIM. No information was avail­able to verify the accuracy of the Tsai-Wu criterion for lam­inated wood structures.

The factors of safety for each model were found at the mast step, daggerboard case and the kingplank at the forestay connection. The results from these analyses are given in Ta­ble 9.

The carbon/epoxy boat showed the highest factors of safety in these three key areas. It is interesting to note that empir­ical construction development resulted in appropriate factors of safety in the st rength.driven daggerboard case design. Fig. ure 11 shows a CQntour plot of the stresses in the laminate material direction in the midsection of a typical IC model.

Conclusions and future developments

The finite element analyses indicated that carbon/epoxy can improve the bending and torsional stiffness, reduce pitch gyradius, and improve factors of safety. Although the fiber­glass model did not perform as well as the all-carhon IC, using FEA to optimize the fiberglass laminates through changes in ply orientation, stacki ng sequence, hybridization, or fiber format (unidirectionals, woven roving, etc.) could re­sult in a better-performing fiberglass IC for lower construc­tion cost than the all-carbon cloth IC. Similarly. optimization or a wood canoe CQuld also produce more competitive boats due to the inherently lower pitch gyradius than the fiberglass boats.

The current state of the art in sailing canoes, as well as other small sai ling craft, is a result of many years of empir­ical development. Unlike the America's Cup where large re­search programs are commonplace, research and develop-

Model

Wood Fiberglass Carbon/epoxy

T.~ 9 Prldletld fK1Of1 of .. Illy

FOS in Mast Step

81 3.1 9.1

ros in Daggerboard C~

21 1.1 2.6

FOS in Kingplank at Fore8tay

6.8 1.0 83

MARINE TECHNOlOGY 303

Ag. 11 COntour plot of stresses It midMction

ment in small craft is generally carried out at the expense of the small boat owner, and teswd in developmental craft like the Ie. With the development of inexpensive analysis tools like those used in this paper, greater analytical study will be available to guide small craft builders, owners and sailors to develop better performing, longer lived, and less expensive craft than those developed empirically .

In the International Canoe, room for improvement still ex­ists in rigs, sails, foils, and hull construction. Experiments with rotating rigs, lighter rig construction, and tougher hulls .... 1.11 continue, and may find their way into mainstream sail­ing as full -batten sails and carbon masts have already. Many other small boat classes will continue to take advantage of the lessons learned in the development classes to improve their perfonnance, increase the enjoyment of their sailors, or lower their costs through careful engineeri ng.

tJS.\

Acknowledgments

The authors would like to thank the following people for their contributions and assistance in the preparation of this paper: Erich Chase, Advanced Marine Composites; Steve Clark, Vanguard Racing Sailboats; Fran DeFaymoreau, Ad­vanced Marine Composites; Ben Fuller. Rockport Appren. ticeshop; Bruce Nelson, Nelson-Marek Yacht Design; Del Olsen, Shore Sails; Ben Robison, Newport News Shipbuild­ing: Steve Slaughter, Maricomp; and Ted Van Dusen, Com­posite Engineering Corp.

References

I Herres.hof, L. F., Th~ CornmDllMtIM' o{Yacht Ift,I#t" , Car.v.n·M.­rine Book •. J.maICA, N.Y .. 1974 , p. 144

2 ~Canoe Salling Rulet," Internatlonal Canoe jo·edt'l"lItion. Jan I.

'98' 3 Stephent, W. p . TrodllwM and MtfnOTU' 0{ Amerlcon Yad/ul,g.

Internatianal Marine Pubhshlng. Camden. Me, 1981, ~. 237 . 4 8~y, M .• Myf/lC 5topai1 MU Kum Wa/~rc:roll Mystic Seaport Mu·

seum, MYl t ic. Conn .. 1979. pp. 189-199. 5 Fo.l . U., Thtl CrUI af/ht Wa~. Charles Scnbnerl Sonl, N.Y .. 1939,

pp. 165-193 6 Fo., U., RIXlng, CrU"II1I. and (k,'IJ". Ashford ?TellS PublIshing,

Shedtitld , Ham~hire. U.K., 1985, pp 27 1- 272. 7 Tht GaUl"''' BroIhu. 0'1 Boat Ca",truc/IO"; Wood aruJ Wt" 5)"$ttm

Mattnall, 4th cd , Gougeon BI'OB. Inc., eay Ci ty. Mich., 1985. 8 Greoene. E. , MonM Composl/tll.· UK 0{ FIMrglo .. " Rtl"for«d PItu ·

tKt," /ht Man"tllfId.u/'7, Report sse 360. Ship Structure Committee. US. Coatt Guard, 1990.

9 Connor, O. and M.the .... L .. ~ Et"ect of the Pro~rtiH af the Con· alltuenf,l on the Fatigue Perfarmance ar CompoeitH: It. Review,~ COM. poIltn, Vol. 20, No. 4. July 1989. pp. 3 17-328

10 Vang. T. Y , F",.k Ekmtlllt 5trw:tural Al\O.l)',u. Prentice·Hall , En· glewood Clif" NJ , 1986, pp. 2-4.

II COSMOSIM UK"" Manual. Vol. 1, ~p 2·1l-2·19, ,nd 2-63-2-66. 12 T .. i, S. W , CornpNlU.lh,'IJ", 4th ed., Think Compoeitn, Dayton,

OhiO. 1988, pp 11 -5-11-6.

Nov~ 7-11,1994 Oftr50Papcn, pb7~"

~ OCTOBER1~

TOpics • SIabiIIty Sd<ty Case -ubop • Cal"i>e nu..bol6-AD SbJp Types and Slza • Stability Informadoo to Masttr

• - SIabiIIty • M.aneuvmng and SllI"riv2llo Storm Seaways • lIuman FactoB in Stability • StabIIlIy Analysb r ... High-Speed Ships

• For additiooal inforlll3llon contxl FIorid2 T reh Stab '94 S«rtUriJt DMsion of Marine and En\'ironmtntal S)'!iltftU Ttkpbonc: US.\---{4(7) 768-8000, exL 8096 Fax U5.-\-( 4(7) 984·8461 , Attn: OE

MARINE TECHNOLOGY